The long term goal is the ability to design more complex, living but artificial neural tissues - a brain on a chip - for use principally as a model for fundamental neuroscience, but, in the process, making scientific progress relevant to neuroprosthetic electrical interfaces, biochip sensors for neuroactive drug screening and toxin testing, and insight into novel biomimetic approaches to computation. The technology should permit investigators novel insights into the nature of communication within neuronal networks, especially as to how the geometric form of the network helps determine its functionality. While the immediate project emphasizes two-dimensional networks, the lessons learned should help in creating and understanding three dimensional networks that will be better biomedical models. While the overall project is risky and perhaps radical, the immediate project is highly conservative in its emphasis on reliability, robustness and repeatability so that, as we begin to use these networks for scientific investigations, scientists are able to repeat the studies and establish statistical significance. This is a neuroengineering project at heart, emphasizing design, test, and system construction. Nonetheless it breaks new ground in the use of surface molecular cues to control multiple cell types and in the application of analytical techniques for the study of neural information transfer. Specific Aim 1: Reliable Geometric Neuronal Circuits, in which cultured neurons follow two-dimensional patterns with high fidelity for long times and with different cellular components. Specific Aim 2: Reliable Functional Neuronal Circuits, in which the stimulus-response patterns are repeatable from trial to trial, day to day, and culture to culture. Specific Aim 3: Neuronal Information Processing, the study of the efficiency with which output neural activity encode input stimuli, with special emphasis on how the spatial configuration of the neuronal circuit affects the encoding, on convergent information flow, and learning paradigms. This research aims to create a "brain on a chip" to permit researchers to better study how the brain processes information. Understanding gained should help improve brain prosthetic electrodes and may lead to new techniques for testing for drugs that may affect the nervous system. [unreadable] [unreadable]